MiMo-V2.5-Pro
Xiaomi · released Apr 27, 2026 · XiaomiMiMo/MiMo-V2.5-Pro
- Type
- Open weightsMIT License
- Params
- 1T
- Context
- 1.1M
about 788K words of context
Our take
Written Aug 3, 2026MiMo-V2.5-Pro is a trillion-parameter text model from Xiaomi with a permissive MIT licence and a one-million-token request limit. It is built for long-document and coding workloads, with measured coding performance well ahead of its general chat score.
Choose this for coding workloads where its measured coding score is the relevant signal, or for long-context text tasks at low entry cost with a genuinely permissive licence. Use it when you want to self-host or modify weights commercially. Skip it if you need image, video or audio support, if creative writing quality matters most, or if you want the cheapest throughput rather than the fastest.
The case for it
- One of the longest request limits among open models we track, at 1,050,000 tokens, with low entry cost.
- Measured coding performance sits 52.6 points above its general chat score — its strongest single benchmark.
- MIT licence permits commercial use, modification and redistribution without restriction.
- Ten hosted offers with a wide throughput range, from 20 to 65 tokens per second.
The case against it
- Creative writing is its weakest measured task, 87.2 points below coding and 34.6 points below general chat.
- Fastest throughput costs nearly three times the output price of the cheapest option, and over four times Xiaomi's own mid-tier offer.
- Text-only: no image, video or audio input or output.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)23rd of 143 · 1465.7
CodingWriting and fixing code on its own
Arena Coding17th of 143 · 1518.3
AgenticPlanning, calling tools, staying on task
Arena Agent (IPS)27th of 36 · −0.025
Arena Agent (IPS) is the only board that has scored it for this.
WritingWe do not rate this
Two boards come close and neither tests writing: Arena Creative Writing asks people which of two replies they prefer, and LiveBench Language tests whether a model understood a passage. So we show where MiMo-V2.5-Pro placed and give it no mark out of five.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done, which is why they get no rating.
Every published score for this model8 scoresEvery figure we hold, from 8 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
- Fits in memory
- weights load entirely on the card
- Spills to system RAM
- some weights offload; much slower
- Too large
- will not load even with offload
- est
- size is calculated; the verdict could change by 10%
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Radeon RX 7900 XT · 20 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Memory use by level
Against a 24 GB card.
All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.
Check against your own machine → · All 71 devices, with every size →
Or rent it from someone else
Cheapest of 10 live listings. Picked at the widest standard context we hold, within one quantisation slice, so the numbers beside it are a price one host actually charges.
- per 1M tokens
- $0.43 in / $0.87 out
- Context served
- 1.1M
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| GMICloudbf16 | $0.35 / $0.70 | 1.1M | 32 tok/s | No | Yesunknown period | Unknown |
| Xiaomifp8 | $0.43 / $0.87 | 1M | 41 tok/s | No | Yes30 days | Unknown |
| AtlasCloudfp8 | $0.43 / $0.87 | 1M | 30 tok/s | No | Yesunknown period | Unknown |
| OpenRouter | $0.43 / $0.87 | 1.1M | not measured | Unknown | Unknown | Unknown |
| Novita AI | $0.48 / $0.96 | 1M | 34 tok/s | No | No | Confirmed |
| StreamLake | $0.52 / $1.04 | 1M | 40 tok/s | No | Yesunknown period | Unknown |
| Novita AI | $0.52 / $1.04 | 1M | not measured | Unknown | Unknown | Unknown |
| DeepInfrafp8 | $1.00 / $3.00 | 1M | not measured | Unknown | Unknown | Unknown |
| DigitalOcean Gradient | $0.60 / $3.00 | 262K | 54 tok/s | No | No | Confirmed |
| DeepInfrafp8 | $1.00 / $3.00 | 1M | 61 tok/s | No | No | Confirmed |
Novita AI
DeepInfra
DigitalOcean GradientAcross the 10 listings we hold: 7 say they do not train on prompts, 0 say they do and 3 do not say. 3 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| GMICloudbf16 | ✗ | ✓ | ✗ |
| Xiaomifp8 | ✓ | ✓ | ✗ |
| AtlasCloudfp8 | ✓ | ✓ | ✗ |
| OpenRouter | ✓ | ✓ | ✓ |
| Novita AI | ✓ | ✓ | ✗ |
| StreamLake | ✓ | ✓ | ✗ |
| Novita AI | |||
| DeepInfrafp8 | |||
| DigitalOcean Gradient | ✓ | ✓ | ✓ |
| DeepInfrafp8 | ✓ | ✓ | ✓ |
Tool calling: 7 of 10 listings say yes, 1 says no, 2 publish no parameter list. JSON output: 8 of 10 listings say yes, 2 publish no parameter list. Strict schema: 3 of 10 listings say yes, 5 say no, 2 publish no parameter list.
Models people weigh against MiMo-V2.5-Pro
When we formed this view
Dates behind this page
Prices last checked 38h ago
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- 2 of 10 listings publish no parameter list, so what their API accepts is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 3 of 10 listings do not say whether they train on prompts.
Licence and identifiers
What the licence allowsMIT License, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
MIT License
Fully permissive: do anything with attribution. No patent grant, unlike Apache-2.0.
Identifiers
- Hugging Face
- XiaomiMiMo/MiMo-V2.5-Pro
- Architecture
- Mixture of experts
- Modality record
- text->text
- Catalogue slug
- xiaomi-mimo-v2-5-pro